INDEX

Explanations

references to significant character actions and motivations within a narrative

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 okuyayım

-0.48

 gethan

-0.42

 kijken

-0.41

 fragen

-0.38

 devriez

-0.37

 tuyệt

-0.37

 ioutil

-0.36

 détaillés

-0.36

 fieltro

-0.35

 Juifs

-0.35

POSITIVE LOGITS

 surla

0.48

 agendas

0.46

 Agenda

0.45

 AGENDA

0.44

 agenda

0.44

 avatars

0.43

########.

0.43

 organisation

0.43

 Organisation

0.42

makeatletter

0.42

Activations Density 0.350%

references to significant character actions and motivations within a narrative

No Comments

No Known Activations